[IPOL discuss] Eigen, once again...
Nicolas Limare
nicolas.limare at cmla.ens-cachan.fr
Sun Dec 18 07:54:14 CET 2011
Hi,
> But I think that support GCC/Visual is a must to have for research code.
I agree that GCC and MSVC are major compilers, and IPOL code should
compile on these compilers. However, this will not always be possible,
because MSVC doesn't implement the C99 standard and there is no plan
to improve the situation[1]. So, C99 code may not compile on the
Microsoft compiler. But C99 is accepted by IPOL.
Actually, IPOL requires (in the recent guidelines) the code to be
standard-compliant C89, C99 or C++98. With these rules, it should
be usable with any compiler correcptly implementing these languages.
[1] http://drdobbs.com/cpp/231900562
> IPOL do not plan to have "company quality code" ...
I do not know what you mean by "company quality"... If you mean
high-performance, finely tuned code, then I agree, IPOL doesn't need
that. IPOL wants "correct readable usable code", nothing more.
> IPOL goal is to provide a usable code that respect an article under some
> basis condition (the basic tools that are used for software development).
... and to do what we can to keep this code usable as long as possible
on as many environment as possible. Nothing can guarantee that a code
will be usable forever. And we have limited means, so we can not test
every configuration. But we can hunt down known sources of
incompatibilities and portability problems. If a non-standard code
is written specifically to compile on GCC and MSVC, it is very
probable that it will fail on other compilers.
> I want to say that we don't have to support all the compiler but the most
> used one.
We are only doing little image processing programs! It is a very
simple software task, we do not touch the hardware, we have no
interaction with the operating system, no network, no GUI, only little
atomic command-line programs. I do not see why we would need more than
what is included by the C/C++ language spec. So I do not see why we
would need to write a code that wouldn't be accepted by any
standard-compliant compiler.
--
Nicolas LIMARE - CMLA - ENS Cachan http://www.cmla.ens-cachan.fr/~limare/
IPOL - image processing on line http://www.ipol.im/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://tools.ipol.im/mailman/archive/discuss/attachments/20111218/2bfa1a59/attachment.pgp>
More information about the discuss
mailing list